Identification of hopanoid biosynthesis genes involved in polymyxin resistance in Burkholderia multivorans

Abstract

A major challenge to clinical therapy of Burkholderia cepacia complex (Bcc) pulmonary infections is their innate resistance to a broad range of antimicrobials, including polycationic agents such as aminoglycosides, polymyxins, and cationic peptides. To identify genetic loci associated with this phenotype, a transposon mutant library was constructed in B. multivorans ATCC 17616 and screened for increased susceptibility to polymyxin B. Compared to the parent strain, mutant 26D7 exhibited 8- and 16-fold increases in susceptibility to polymyxin B and colistin, respectively. Genetic analysis of mutant 26D7 indicated that the transposon inserted into open reading frame (ORF) Bmul_2133, part of a putative hopanoid biosynthesis gene cluster. A strain with a mutation in another ORF in this cluster, Bmul_2134, was constructed and named RMI19. Mutant RMI19 also had increased polymyxin susceptibility. Hopanoids are analogues of eukaryotic sterols involved in membrane stability and barrier function. Strains with mutations in Bmul_2133 and Bmul_2134 showed increased permeability to 1-N-phenylnaphthylamine in the presence of increasing concentrations of polymyxin, suggesting that the putative hopanoid biosynthesis genes are involved in stabilizing outer membrane permeability, contributing to polymyxin resistance. Results from a dansyl-polymyxin binding assay demonstrated that polymyxin B does not bind well to the parent or mutant strains, suggesting that Bmul_2133 and Bmul_2134 contribute to polymyxin B resistance by a mechanism that is independent of lipopolysaccharide (LPS) binding. Through this work, we propose a role for hopanoid biosynthesis as part of the multiple antimicrobial resistance phenotype in Bcc bacteria.

Fig 1

Map of the B. multivorans ATCC 17616 locus affected by Tn5-751S in mutant 26D7. The region from base pairs 2344808 to 2350836 of the B. multivorans ATCC 17616 genome sequence (http://genome.jgi-psf.org/burmu/burmu.home.html) is shown in the published orientation with ORFs Bmul_2132, Bmul_2133, and Bmul_2134. The locations of the transposon insertion in the 26D7 mutant and key restriction sites are marked. The DNA fragments used for complementation experiments are indicated by the black bars above the sequence. The region of Bmul_2134 deleted in the RMI19 mutant construct is indicated by the black bar in the ORF. A, AfeI; Ap, ApaI; B, BamHI; N, NotI; S, SalI; P, PstI.

Fig 2

Effect of increasing concentrations of polymyxin B on the outer membrane permeability of B. multivorans ATCC 17616 compared to mutants 26D7 and RMI19. Polymyxin B was titrated into suspensions of whole bacteria at an OD₆₀₀ of 0.5 and 10 μM NPN. An increase in fluorescence corresponded to an increase in membrane permeability and uptake of NPN into the hydrophobic periplasm. Data were calculated as the fluorescence with polymyxin B minus the fluorescence without it and represent means for three biological replicates ± standard errors of the means. (A) Mutants 26D7 and RMΙ19 were each significantly more permeable to NPN than the parent strain in 0.78 to 200 μg/ml of polymyxin B (P < 0.02; t test). ■, B. multivorans ATCC 17616; □, B. multivorans 26D7; ◊;, B. multivorans RMI19; ▴, P. aeruginosa ATCC 27853. (B) Outer membrane permeability of mutant strains in the presence of 100 mg/ml polymyxin B is restored when the mutations are complemented in trans. P. aeruginosa ATCC 27853 was used as a positive control. B. multivorans 26D7(pUCP26) and B. multivorans RMI19(pUCP28T) were significantly more permeable to NPN at 100 mg/ml polymyxin B than the B. multivorans ATCC 17616 parent strain with the respective empty vector (P < 0.05; ANOVA). B. multivorans 26D7(pRM308) was not significantly different from the parent or mutant strain with pUCP26. The permeability of B. multivorans RMI19(pRM305) to NPN was significantly less than that of RMI19(pUCP28T) at 100 mg/ml of polymyxin B (P < 0.05; ANOVA).

Fig 3

Binding of DPX to whole parent and mutant cells. Bacterial cells were treated with increasing concentrations of DPX, and fluorescence was determined by emission at 485 nm upon excitation at 340 nm. ■, B. multivorans ATCC 17616; □, B. multivorans 26D7; ◊, B. multivorans RMI19; ▴, P. aeruginosa ATCC 27853. The background fluorescence of DPX with buffer alone was subtracted at each concentration. The data represent the means for three biological replicates ± standard errors of the means.

Fig 4

Neutrophil bacterial killing by cells from a patient with CGD. CGD neutrophils were incubated with bacteria (1:1) at 37°C. Viable bacterial counts were determined at 0, 30, and 120 min postinfection. Data presented are percentages of bacteria relative to the starting inoculum at each time point. ■, B. multivorans ATCC 17616; □, B. multivorans 26D7; ●, B. multivorans JTC; ▴, P. aeruginosa M2 as a susceptible control (43).